Phase Space AI

Valuation

Lumentum Holdings [LITE]

Lumentum Holdings [LITE] — Valuation

Spot $602.23 (2026-07-29). Two outputs, as required: a 12-month target and the 5-year implied-path test.

Read §2 before anything else. LITE's fully diluted share count is 102.1m against 77.8m common outstanding, and every valuation number here rests on that.

---

1. COMPANY STATE — declared first

STATE C — scaling but economically observable

Test, applied:

Not A: margin variance is enormous (GAAP quarterly operating margin −43.3% to +21.7% across the window) and there are three structural regime breaks inside eighteen months — the Cloud Light acquisition (Nov 2023), the retirement of segment reporting (Q1 FY2026), and the $2.0bn NVIDIA preferred plus $1,265m of 2032 Notes (Sep 2025 / Mar 2026). Not B: LITE's driver is not an exogenous commodity cycle but a product transition it participates in. Not D: the reverse DCF solves across the entire range; the instrument applies.

Consequence of State C, per valuation.md

• Terminal margin is BUILT, never assumed, through m_EBIT,T = m_gross,T − R&D − S&M − G&A − other.
• The hard arithmetic constraint m_EBIT,T ≤ m_gross,T fails the model rather than flagging it.
• The ceiling is necessary but not sufficient — the full expense bridge is the real test.
• Instrument: the two-dimensional expectations surface {(g, m) : V(g,m) = P₀}. Report all three: required margin at the underwritten growth, required growth at the underwritten margin, and the region clearing the hurdle. §4 does this.

evidence_grade: C. This is the lowest of the three names and the reasons are specific: the segment disclosure was retired after a 92% profit decline in the affected segment; the ≥10% customers are not named; design wins are not disclosed; consensus is unavailable; the fully diluted share count had to be reconstructed from three separate filing notes; and the AV data for this name carries eight distinct defects including duplicated quarters. The current-period financials are solid and EDGAR-tied. The disclosure regime is the weakest in the cluster.

---

2. THE SHARE COUNT — the largest single input, and no vendor field carries it

All four note series are deeply in the money — 2026 at $99.29, 2028 at $131.03, 2029 at $69.54, 2032 at $187.77, against a $602.23 share price. Principal $3,198.4m; estimated fair value $19,074.7m. LITE settles principal in cash and the $15.9bn excess in shares.

Independently confirmed against a source the vendor did not produce. CFO, 2026-05-05: "These projections also assume shares used for non-GAAP diluted earnings of approximately 102 million shares." The reconstruction ties to 0.1%.

What using the wrong count does

The naive count understates the required growth rate by 8.2pp and enterprise value by 31.5%. AV returns 96.2m for this field; the screen returned nothing at all.

One offset declared rather than netted: LITE paid $102.0m for 2032 Capped Call Options, which economically offset 2032 dilution up to a cap that was not extracted within the time box. Capped calls do not enter GAAP diluted EPS. The $61,613m EV is therefore the conservative end of a $57.6–61.6bn range, and it is used because it ties to the company's own guided count.

Enterprise value: 102.10m × $602.23 + net debt $124.5m = $61,613m. Net debt: convert principal $3,198.4m + term loans $98.4m − cash $2,617.8m − short-term investments $554.5m = $124.5m, hand-totalled from the 10-Q debt note. Operating leases excluded, per the convention stated across all three names in this cluster.

Current multiples: EV/Sales 24.76x · EV/EBIT 245.2x · EV/EBITDA 121.8x.

---

3. TERMINAL MARGIN — 28.0%, built, and unprecedented-and-verified

Underwritten terminal EBIT margin: 28.0%.

The bridge

Hard constraint m_EBIT,T ≤ m_gross,T: 28.0% ≤ 47.0% ✓. And the sufficiency test the constraint alone does not cover: the bridge leaves 19.0pp of revenue for all R&D, selling and administrative cost, against 39.7pp in FY2025. On $4.9bn of revenue that is $931m of absolute opex against FY2025's $652m — a 43% increase in absolute spend. The bridge is affordable in absolute dollars, which is the test the record shows being skipped (one name carried a 14.4% terminal margin against an 11.9% gross margin).

Why 28.0% is far above the trailing TTM GAAP actual of 10.10%

valuation.md requires an explicit causal bridge for any terminal margin materially away from trailing. Here it is, and it is corroborated rather than asserted:

1. The TTM window is contaminated by two pre-inflection quarters. Q4FY25 printed −0.7% and Q1FY26 +2.8%. The exit rate is +21.7% (Q3FY26 GAAP), and the guided Q4FY26 non-GAAP figure is 35–36%. Anchoring on the TTM average would anchor on a period that no longer exists.
2. The company guides above the underwritten figure. 28.0% GAAP against 35–36% non-GAAP is a 7–8pp haircut for SBC and residual amortisation. The terminal margin sits below management's own next-quarter guide, which is the correct direction of conservatism for a State-C name.
3. The operating leverage is arithmetic. Revenue per quarter went $425.2m → $808.4m in four quarters while R&D was approximately flat in FY2025 versus FY2024 ($303.9m vs $302.2m). Fixed-cost absorption, not margin optimism.

Unprecedented-and-verified, declared explicitly

LITE has never earned a 28% operating margin. Its GAAP operating margin was −43.3% as recently as FY2024. Per the brief's operative distinction, this is historical implausibility, not arithmetic impossibility, and the rule is: verify against a second source and, if it holds, adopt it and say so.

• Source 1: the Q3FY26 10-Q — GAAP gross margin 44.2%, GAAP operating margin 21.6%.
• Source 2: the 2026-05-05 earnings call — non-GAAP gross margin 47.9%, Q4 non-GAAP operating margin guided 35–36%, non-GAAP EPS $2.85–3.05.
• Source 3: av_vs_edgar.py — 0 DISAGREE on the periods compared.

Adopted, and flagged as unprecedented-and-verified. A bound calibrated on LITE's FY2024 trough would reject the single largest margin inflection in this cluster's history, which is exactly the error the brief spends a full section warning against.

No screen figure to compare against: LITE was returned INDETERMINATE and carried no terminal margin at all.

---

4. IMPLIED-PATH TEST — the Valuation Criteria

Terminal value exceeds 95% of EV, so the reverse DCF is the mandatory primary long-horizon output.

reverse_dcf.py --spot 602.23 --shares 102.10 --net-cash -124.5 --revenue 2488.4
               --years 5 --wacc 0.10 --terminal-margin 0.280 --solve cagr

Held fixed and named: terminal EBIT margin 28.0%; WACC 10.0%; horizon 5 years; TTM revenue $2,488.4m; fully diluted shares 102.10m (derived and confirmed against the guided count); net debt $124.5m; EV $61,613m.

The exit multiple, derived first

EV_T / EBIT_T  =  (1 − t)(1 − g/ROIC) / (WACC − g)

Base exit multiple adopted: 19.0x — above the 16.1x identity base, and the reason is argued below rather than applied silently.

Reference-class assertion: peer-derived exit multiple is UNIDENTIFIED

valuation.md requires that an exit multiple be drawn only from a comparator set whose growth brackets the subject's growth at the exit year, with n≥5 mature profitable firms and a matching operating model.

The class fails on two counts and the multiple is declared UNIDENTIFIED:

1. No comparator brackets LITE's exit-year economics. The names whose growth brackets LITE (COHR, CIEN, MRVL) have GAAP EBIT so depressed by acquired-intangible amortisation that their EV/EBIT multiples — 65.2x, 78.1x, 104.2x — carry no information about a warranted continuing multiple. The names with clean high multiples (CRDO 73.4x, ALAB 196.6x) grow 2–7x faster than LITE and are fabless, not integrated manufacturers.
2. This cluster reproduces the §9.9 failure directly. EV/EBIT ranges from 24.1x to 245.2x across comparators whose operating margins range 10.1% to 43.7% — the multiple is tracking amortisation accounting, not growth. A multiple drawn from it would be a confident number with no discriminating power, which the reference calls worse than an admitted UNIDENTIFIED because it looks like an answer.

So the base 19.0x is derived from the identity, not from peers. The argument for 19.0x over the 16.1x identity base — required because a base above a derived anchor needs one as much as a base below: LITE's terminal ROIC is plausibly higher than the 28% assumed (it operates near-zero net debt, its incremental capacity converts a GAAP fab rather than building one, and its acquired-intangible amortisation is contractually finite), and at ROIC 30% with g 7% the identity yields 21.0x. 19.0x sits between the base and sustained parameterisations.

The required parameter

Flip point against TTM YoY: exit multiple 10.3x. Against the 3-year CAGR the test fails at every multiple below roughly 60x.

The margin the strategy ranks on

Demonstrated − required = 69.0% − 49.61% = +19.39pp.

The other two dimensions of the surface, as State C requires

Required margin at the underwritten growth:

The (g, m) region clearing the hurdle: at a 19.0x exit and a 28.0% terminal margin, the price is cleared by any pair with g ≥ 49.6%; at g = 45% a 32.6% terminal margin is needed; at g = 40%, 39.0%. The surface is passable only in its top-left corner — very high growth, or a terminal margin above anything in the sector.

The implied compression, as a number

Current EV/EBIT 245.2x → base exit 19.0x: a compression of 226 turns, 92.3%. On EV/Sales, 24.76x → an implied 19.0x × 28.0% = 5.32x, a 78.5% compression. This is by far the largest implied compression of the three names.

Verdict: PASS WITH ARGUMENT — with a recorded dissent

On the letter of criteria.md, this is a PASS. The rule is "PASS — the implied path sits at or below what the business has already demonstrated," and 49.61% required sits 19.39pp below the 69.0% demonstrated. LITE is the only one of the three names to clear on the primary measure.

The argument is specific, evidenced and quantified, not narrative: - Q4FY26 guided $985m midpoint — an all-time record and a +104% YoY rate. - Non-GAAP operating margin guided 35–36%, against 20–22% guided two quarters earlier. - Three named growth drivers with the magnitude explicitly still ahead: cloud transceivers, OCS, CPO. "the vast majority of this growth is still ahead of us" (CEO). - OCS is already in revenue and quantified: >$38.0m in 9M FY2026 — the only quantified new-product revenue figure in this cluster. - Cloud transceivers grew by more than $268.0m over the same nine months (10-Q). - 1.6T margins "significantly better than 800G" (CFO, under Q&A).

The dissent, recorded rather than buried, because it is the more important half:

1. The required path is a LEVEL claim, and the level is not supported by anything LITE discloses. 49.61% for five years takes revenue from $2,488m to $18,652m by FY2031 — larger than COHR, CIEN and LITE combined today. No backlog figure, no capacity plan and no TAM number in LITE's filings supports it. The company's own stated ambition is a $2 billion quarterly run-rate — $8bn annually, less than half the required level, and with no date attached.
2. The PASS is an artefact of measuring "demonstrated" over 12 months. A one-year rate of 69% cannot demonstrate a five-year path. On the three-year CAGR from filed annuals (19.2%) the test fails by −30.41pp. On any sustained rate a reasonable person would underwrite — 40%, itself extraordinary — the margin axis fails by −11.02pp.
3. The multiple is at ≥91st percentile of its own history (§5) and the required compression is 92.3%.

Stated plainly: LITE passes the Valuation Criteria as written, and it is the name in this cluster where the criterion's construction is doing the most work. A book that ranks on demonstrated − required should be told that LITE's +19.39pp becomes −11.02pp the moment the growth window is extended past twelve months, and that this is the single most fragile number in the three-memo set.

Sensitivity — on the exit multiple, never on scenario probabilities

The table above is the sensitivity. A 1.0x change in the exit multiple moves the required CAGR by ~1.6pp at the base. But the highest-variance parameter on this name is not the exit multiple — it is the growth window used to define "demonstrated", which moves the verdict by 49.8pp between the 12-month and 3-year measures. That is disclosed here rather than resolved, because resolving it is a strategy decision.

---

5. TWELVE-MONTH TARGET

Target $827. +37.3% to spot $602.23.

valuation.md step 1 requires near-term consensus. AV EARNINGS_ESTIMATES returns an empty array for LITE (and COHR and FN) while returning 22–41 rows for CRDO, CIEN and AAOI on the same premium key in the same minute. The substitution is declared: filed guidance replaces consensus.

Build

The reflexive share count — iterated, not ignored

LITE's diluted count is a function of the share price, because the convert excess-value settlement grows as the stock rises. A target computed on today's 102.10m count is internally inconsistent.

Converges at $827, +37.3%. The naive single-pass answer is $839; the iterated answer is $827. A 1.4pp difference, recorded because no screen or vendor field performs this iteration and because the direction is always adverse to the holder.

Named product-cycle events inside 12 months

Dated in LITE_Catalyst_Calendar.md: Q4/FY2026 results and the first FY2027 guide (~mid-August 2026); the FY2026 10-K — the first annual report filed under single-segment reporting, and therefore the first that will not disclose the Industrial Tech split; LITE's own 1.6T transceiver shipments from mid-CY2026; the 102.4T switch-silicon gate in late Q2 CY2026 that the industry 1.6T ramp waits on; continued 2026 Note conversions; and CPO's first material revenue.

Caveat, and it is larger here than on the other two names

Three things all have to hold: revenue must roughly double again in FY2027, the multiple must stay near the 80th percentile of its own five-year range, and the diluted count must not grow beyond the iteration above. At the five-year median multiple the target falls roughly 65% below spot. LITE has the highest upside of the three names and by far the widest distribution around it.

No external professional target was available for the sanity band. Recorded as absent.

---

6. Liquidity Criteria and vehicle — PASS, equity

Vehicle: EQUITY. Ladder rung 1. No argument required, and none is offered.

Why not a LEAP. The implied-over-realised premium is the narrowest of the three names (+7.1 to +10.4 points against COHR's +12.5 to +14.0), so on price alone LITE is the most defensible option candidate in the cluster. It fails on depth instead. The $600 July-2027 strike quotes 10 × 13 contracts against 29 of open interest — the thinnest quote of any strike checked across the three names, and squarely in the territory the framework's own record calls uninvestable (HCA 18 contracts, GMED 13, CRDO 78). A $4 bid-ask on a $230 premium is 1.7%, which looks fine, but 10 contracts of size means the quote is decorative.

Chain depth is not strike depth, and this is a clean instance: LITE's chain looks reasonable in aggregate and every individual strike near the money is 6–30 contracts. Equity.

Additional reason specific to this name: LITE's equity already embeds substantial convexity through the convert overhang — the diluted share count rises with the price, damping upside, and the $3.2bn of cash-settled principal is a fixed claim. Buying an option on top of a capital structure that is already option-laden compounds a risk the sizing rule does not capture.

---

7. MEASURED Criteria

---

8. Summary card